You are on page 1of 30

A Regional Network Strategy for Reciprocal Rural-Urban Linkages:

An Agenda for Policy Research with Reference to Indonesia

Mike Douglass

Cecilia Tacoli, ed. (2006), Rural-Urban Linkages (Earthscan), 124-154.

Reprint from Third World Planning Review, Vol 20, No. 1, 1998

Abstract. A regional network strategy incorporating rural-urban linkages and interdependecies in its analysis and
policy framework is proposed to overcome the continuing rural–urban divide in development theory and planning.
The strategy is illustrated by drawing from variations in regional contexts in Indonesia. A comparison of a virtuous
cycle with a truncated process of rural regional development is used to set an agenda for policy research. Research
on five types of rural-urban flows -- people, production, commodities, capital, and information -- is needed to
provide baseline data to align policies with a more reciprocal relationship between rural and urban development.

1. Overcoming the Urban and Rural Divide in Planning

Debates on the nature of rural-urban relations hold a prominent position in development theory
and planning. Discussions in the 1950s centered on whether towns played parasitic or generative
roles in their relations with their rural hinterlands. As originally argued, generative forces of
modernization associated with urbanization were hypothesized to outweigh possible parasitic
impacts on rural areas (Singer 1964). The Lewis model of economic development widely
adopted at the time also accepted the idea that economic growth and modernization required a
transfer of surplus from an assumed moribund agricultural sector to urban-industry, thus
justifying the appropriation of rural resources, labor and capital by cities in the name of longer-
term national economic development (Fei and Ranis 1964).

The general policy prescription derived from the Lewis and related models called for an
acceleration of urban-industrial growth and the transition to an urban-based society. The concern
raised in many Asian countries up to the 1970s was that levels of urbanization were very low and
that the push of rural poverty was causing rural-urban migration to create an involuted, distorted
urban growth process based on the expansion of low-productivity urban services rather than on
the dynamics of industrialization. Whether from a modernization or economic perspective, the
prevailing view was that increasing the pace of a more “authentic” pattern of urban-industrial
development was needed to compensate for inherent limitations in rural development.

From the late 1950s, an opposing view emerged in a new field of regional planning that was
built, in part, on core-periphery and spatial polarization models, which observed that in most
developing countries the benefits of economic growth increasingly concentrate in one or a few
core urban regions. The principal thesis of these models was that the benefits accruing to the
core were at the expense of the rural periphery. Cities organized rural areas to serve urban
interests, resulting in net capital outflow, brain drain and other resource transfers that lowered
rather than raised the potential for rural areas to develop. Cities actively exploited rural areas,
with rural poverty and rural-urban migration not emanating from a the isolation of rural from
urban areas, but rather from the tightening of rural and urban linkages. The further observation
that rural areas were often transformed into overly-specialized single-crop or natural resource
economies to serve urban-based corporate interest fit into the parallel emergence of dependency
theory, which argued that the metropolitan nations of the North actively under-developed the
agrarian economies of the South. Rural-urban linkages were thus part of global chains of power
and control the perpetuated conditions of rural poverty and underdevelopment.

Although the idea that cities act as “theaters of accumulation” inhibiting rather than fostering the
development of rural regions (Armstrong and McGee 1985) has continued to find adherents, the
position that gained wider agreement was that the “backwash” effects of urbanization on rural
areas were expected to be only short-term, i.e., occurring only in the “early stages of
development” (Williamson 1965). As spatial systems matured, development impulses were
expected to naturally become more readily articulated over national territories, thus reversing
previously dominant polarization processes.

Although much research was stimulated by this thesis, it did not result in any consistent evidence
that polarization reversal has either occurred in a sustained manner in any developing country or
that it is an automatic outcome of increasing levels of per capita GDP reaching "mature" stages
of development (Douglass 1990). The general case could still be made, therefore, that if a
polarization reversal were to occur, it would be best pursued through government interventions.
The question of what kind of policy interventions would be needed was given a single answer:
induced urbanization in the periphery, which came to be known as the growth pole or growth
center approach to regional development (Friedmann 1968). As a strategy, it argued that only
the rise of cities in the periphery could challenge the growth of core urban centers. As such, it
has no explicit rural content; rural areas are still seen as being developed by the city and remain
“out there” as uninvestigated hinterlands. The absence of any rural development component in
this strategy is glaring, but despite this and many other criticisms of the growth pole model
(Friedmann and Weaver 1979), notably those positing that in a transnational age the likelihood
that linkage and spread effects will be captured locally is extremely low, this remains the
dominant spatial development strategy throughout the world, including Asia.

In the 1970s a new slant on the idea that cities were the cause rather than solution to rural
problems received a new thrust under the title of “urban bias” in rural development, which
argued that rural underdevelopment is perpetuated by unrelenting political, social and economic
forces favoring cities over the countryside (Lipton 1977). Subsequent treatments of the same
theme cited the development planner as a principal culprit in putting rural development last
rather than first. The incentive to be in the capital city for career advancement, the low status of
agricultural economics, the avoidance of rural visits during the rainy season when conditions
were worst, the disdain of actually staying overnight in a village, and the diplomacy of not
asking rural elites embarrassing questions about social cleavages or poorer village residents were
part of the systematic biases preventing planners from comprehending rural issues and taking
appropriate actions (Chambers 1985).

While most of the debates on rural-urban relations remained abstract and without conclusive
empirical evidence supporting either pro- or anti-urban stances, over the decades the cumulative
result has been a curious divide in planning. On one side are urban planners who continue to
plan for rural development from cities and who know exceptionally little about and give scant
regard to the potential for an agriculture- or rural-led development. Their rural landscapes
consist of urban nodes and transportation linkages overlaid on topographical maps. For them

regional integration through urbanization remains the key. Their policies have a decisive urban

On the other side are rural development planners who tend to view cities as parasitic and alien to
rural interests. Wary of cities, they rarely put them in their rural landscapes, preferring instead to
define rural areas as consisting only of agricultural plots, resource areas and villages. Efforts at
integrated rural planning generally use the village and its agricultural land as the highest units of
development. The resulting policies for typically have a discernible rural bias with little or no
interest in investigating how cities might be better brought into rural planning frameworks.

To a rural household, however, the landscape of daily life includes both rural and urban
elements. Rural-urban linkages are part of the local reality for household members carrying out
the diverse tasks of producing income on and off-farm, maintaining a living space in the village,
and going to local and even distant towns for shopping, marketing, work, and specialized
services. The challenge for rural regional planning is to overcome the rural-urban divide by
incorporating this reality into development frameworks and, further, identifying policy measures
to foster mutual benefits for both town and village households.

The purpose of this paper is to sketch the outlines of regional planning framework that
incorporates rural-urban linkages and can adjust to a variety of local situations. The reference
for this discussion is Indonesia. Section 2 focuses directly on rural-urban relations and argues
the case for viewing rural and urban development as being interdependent. Section 3 places the
question of rural-urban linkages into a larger regional network framework that advocates the
organization of planning around regional networks (clusters) of rural and urban linkages.
Section 4 illustrates the regional network model through examples of types of situations in
Indonesia. Section 5 concludes with an agenda for policy research.

2. Rural-Urban Linkages and Interdependencies

Formulating comprehensive rural-urban development frameworks to explicitly promote rural

development began to receive attention in the 1970s and 1980s. Some attempts, such as the
“urban functions in rural development” approach discussed below, were thinly disguised forms
of urban bias. Others, such as the many programs put forth under the rubric of “integrated rural
development” (IRD), erred in the opposite manner by rarely including explicit urban
components. While making the important contribution that rural development must be seen as
being multi-faceted, including non-agricultural as well as agricultural activities, IRD was
essentially aimed at improving planning coordination through integrated sectoral planning in
specific localities (Cohen 1980; Birgegård 1988). Although activities continue to be organized
under the IRD logo, even where the spatial dimension is implied it is almost always limited to
marketing functions; the tendency to treat rural as a village-agriculture question continues.

Among the few frameworks put forth that explicitly united rural with urban development was the
“agropolitan” approach (Friedmann and Douglass 1978; Friedmann and Weaver 1979; Douglass
1981). It proposed that rural development could be best pursued by linking rural with urban
development at the local level. Setting an agropolitan development process in motion required
attention to at least three critical issues: access to agricultural land and water, devolution of
political and administrative authority to the local level, and a shift of national development

policies in support of diversified agricultural production. Seeing the rural town as a principal site
for rural non-agriculture as well as political-administrative functions rather than as an industrial
growth pole, the agropolitan approach suggested that in most countries the district scale was
most appropriate unit of development in that it was small enough to afford frequent access to
urban functions by rural households yet large enough to expand the scope of economic growth
and diversification to overcome the limitations of using the village as an economic unit. It also
proposed that local knowledge could better be incorporated into planning processes if it were
organized at a level that was close to rural producers and households.

While the agropolitan approach has, at best, been only partially applied in practice, its promotion
of decentralization and democratization has emerged as the most vital issues in planning in Asia
today. The challenge suggested by the agropolitan approach and more fully developed elsewhere
(Friedmann 1992) is how to incorporate local capacity building and popular participation into a
program to foster mutual benefits for rural and urban areas in the course of national

2a. The one-sided view of the “role of cities” in rural development

Although development planning continues to be marked by a divide between rural and urban
specialists, over the years a minor tradition focusing on the role of cities in rural development
has emerged. In the 1970s the idea that cities could assist in promoting development in rural
regions began to focus attention on rural towns at the bottom rather than top of the urban system.
With roots in central place theory, a seminal work in this area was that of Johnson (1970), who
argued from the case of India that the development of rural towns to fill in the gap between the
“parasitic big city” and the village was a necessary condition for the commercialization of
agriculture. The task of the planner was thus one of promoting the growth of market towns.
With central place theory was the overall framework, towns also had to be identified in relation
to each other to ensure appropriate spacing and functional composition by size and position in
the urban hierarchy.

Subsequently embellished by Rondinelli as the "urban functions in rural development approach"

(UFRD), Johnson's advocacy of selecting key rural towns as rural development centers was re-
interpreted as a set of rural development roles to be played by towns and cities n rural regions.
Among the roles identified (Rondinelli 1979, de Jong 1988) were:

(1) consumer convenience centers for purchasing non-durable and durable goods
(2) centers for higher order public and private services
(3) linkage to (inter-)national markets for selling rural products
(4) production supply and support centers
(5) agro- and resource-processing centers
(6) non-agricultural employment for rural labor
(7) centers of information and knowledge

This listing of roles has received continuing interest among regional planners. The roles have,
however, proven to be difficult to translate from abstract generalizations to either real world
experiences or policy recommendations. Several reasons account for this difficulty. First, the
generalizations are very high off the ground and offer no straightforward way of accounting for

the rich variety of (rural) regional contexts within which cities play their developmental roles.
The role that a city plays in a rice growing region, for example, can be expected to be quite
different than that at the fringe of an industrializing metropolitan region. Thus although the list
of potential roles is intuitively appealing, it does not by itself indicate how each roles is expected
to be manifested in specific contexts and, therefore, how functions should be promoted in
different cities.

Second, methods for selecting key rural towns for investments are not clear. Should towns
already playing these roles be selected, or should towns with potential for playing these roles be
chosen and, if so, how is “potential” to be defined? Since most of rural Asia has experienced
relatively slow rates of economic growth, and since rural towns have not been found to be
playing the many roles assumed for them, if towns are to be used for a national program of rural
development, identifying those with high potential becomes a central task. But because such
potential can only be defined in relation to the opportunities and needs of each rural region, it
follows that the potential role of each rural town or city must arise from the regional context
rather than from the study of a town alone. However, the methods developed for selecting key
cities focused only on the attributes of the towns themselves, and in the only real world
application in the Bicol Region of the Philippines, no study of rural potential was included in the
policy research program (Rondinelli 1979; Koppel 1987).

As an example of the severe conceptual and practical problems encountered in mechanistically

using central place theory to construct a rigid set of functions for all towns at a given level of the
urban hierarchy, a study of rural towns by Cohen et al. (1977) showed how town functions
varied among four types of regions in Malaysia, i.e., regions dominated by the kampung
(smallholder rice producers) sector, the estate sector, new town settlements in frontier areas, or
extensive (forestry) sector. The study found that the assumption that “town size and urban
functional diversity are positively correlated does not hold for those centers which lie at the base
of the urban system in developing countries” (p. 12), and further that (p. 23), “The organization
of the regional urban space is a function of the form and structure of agriculture -- and not the
other way around,” with differences in agricultural crop regimes and the organizational structure
of production being crucial determinants of variation in the functional make-up of rural towns.

Third, and even more problematic, is the assumption that all urban functions are developmental
and none has negative impacts on rural areas that outweigh positive ones. Whether the function
is that of a bank, a real estate agent, a timber company office, or administrative office, if it is
located in a rural town, it is assumed to benefit rural hinterlands and all rural households
irrespective of social or economic class. In applications of the UFRD approach in the
Philippines, it was assumed, for example, that locating a bank in a town would automatically
promote rural development by moving formal financial institutions and credit closer to rural
areas. Yet, as studies in Thailand have shown, although the majority of bank branches are
located in rural areas, the agricultural sector typically receives an extremely small portion of
their loans.

The point to be drawn from this observation is that programs to foster the growth of urban
functions for rural development require specific attention to their actual rural linkages and
impacts rather than assuming them to be relevant and beneficial in any and all circumstances.
Even where urban functions may reach rural households, they may do so in such an uneven

manner that they heighten rather than ameliorate problems of rural inequality and poverty. In
situations of unequal distribution of land, for example, subsidized government credit institutions
set up in rural towns for agricultural development have been found to be biased toward rural
elites and large farmers who have channeled loans into further land purchases, leading to
increased tenancy rates and widening rural income disparities (Khan 1979). This is the general
case made by Lipton (1977) with regard to urban bias, namely, that urban elites co-opt or give
patronage to rural elites as a means of facilitating rural-to-urban surplus transfers, with the net
result that urban poverty is perpetuated by the extension of the government into rural
development programs. To put the matter directly: urban functions are not socially neutral.
Moving an activity or institution to a rural town may be an important first step toward promoting
rural development, but care must be taken to link these functions to the specific tools and means
of increasing the spatial and social access by targeted rural households.

2b. Research on the actual roles of towns in rural development

Research on actual roles played by selected towns and growth centers in rural regions in the
Philippines, Malaysia, Thailand and Indonesia reveals the limitations of attempting to stimulate
rural development solely through promoting the growth of towns. One of the earliest efforts to
use selective investments in rural towns as a means to promote rural development was the
USAID funded "urban functions in rural development" (UFRD) approach noted above. Carried
out in the Bicol Region in the late 1970s, the exercise was part of a grand Tennessee Valley
Authority-like river basin development scheme that incorporated dam and irrigation development
with agricultural projects and urban development into a comprehensive regional development
program. Administered under a special Bicol River Basin Development Authority headquartered
in Manila, this ambitious program was meant to turn one of the Philippines poorest regions into
one of its most dynamic and, at the same time, undercut anti-government insurgency in the

The program confronted problems typically associated with central planning, namely, poor
coordination among ministries, and was promoted and carried out without any proposal for land
reform and other institutional changes either before or during the planning process.
Decentralization and participation were given token regard. Instead, the approach taken by the
UFRD team was to fill in the supposed missing functions in rural towns to stimulate rural
development. No study was done of the conditions or needs of rural hinterlands. A study
conducted 10 years after its initiation found that not only were the towns not performing the
"missing" functions, but that they were themselves stagnating (Koppel 1988).

Some of the contributing reasons for the general failure of this attempt offer constructive thought
for rethinking rural-urban relations and provincial urban growth. First, identified functions (e.g.,
barber shops, cabarets and other entertainment) did not necessarily have much to do with rural
development and were, in fact, supported by an unusual presence of national military and civil
service personnel in the region. Secondly, transportation development linking this region to
larger towns outside of the region usurped rather than encouraged the expansion of marketing
and commercial functions of the local towns. Third, since an advancement of agricultural failed
to materialize, the principal source of the growth of central places, i.e., rising rural household
expenditures for non-agricultural goods and services, also failed to emerge to generate the
expansion of towns.

In a related study in Indonesia, three roles were assessed for the Central Java town of
Banjarnegara: (1) consumer shopping and service center; (2) antipode to rural-urban migration
to metropolitan regions (Jakarta); and (3) support and marketing center for agriculture (de Jong
et al. 1983; de Jong 1988). The investigation found very limited performances in all roles.
Concerning (1), consumer shopping and services, a large portion of the district population made
no use of the town for shopping, education, health or bank credit services. As a generator of
employment alternatives (2), job expansion was, like the town itself, found to be very slow in a
region already plagued by high levels of rural underemployment. Finally, as an agricultural
support center (3), although the town was an important node in the marketing chain for fruits and
cattle, the more important rural production activities in coconut and vegetables bypassed the
town entirely. Furthermore, many of the buyers for fruit and cattle were not town dwellers, but
came from outside the region on marketing days, thereby limiting the local employment and
related multiplier effects that could have been captured by the town. A similar study conducted
in a Malaysian town, Pasir Mas, located along the Malaysia-Thai border drew parallel
conclusions (Maude 1983).

Whether such studies are exceptions to a generally more positive of rural towns is an important
question in need of further research. Unfortunately, few studies have been carried out on the
actual relationships between town functions and rural access to or benefits from them. However,
indirect evidence suggested by the generally slow growth of rural towns and persistence of
poverty in many rural regions of Asian countries indicates that rural towns have yet to play many
of the hoped-for roles assigned to them. Studies carried out in Asia, Latin America and Africa in
the early 1980s found that towns in densely-settled rural regions located away from metropolitan
and frontier regions were growing at rates roughly equal to natural population growth, indicating
that employment creation for their hinterland labor force has been close to nil in most cases
(Mathur 1982; Kammeier 1984).

Whether such studies are exceptions to a generally more positive of rural towns is an important
question in need of further research. Unfortunately, few studies have been carried out on the
actual relationships between town functions and rural access to or benefits from them. However,
indirect evidence suggested by the generally slow growth of rural towns and persistence of
poverty in many rural regions of Asian countries indicates that rural towns have yet to play many
of the hoped-for roles assigned to them. Studies carried out in Asia, Latin America and Africa in
the early 1980s found that towns in densely-settled rural regions located away from metropolitan
and frontier regions were growing at rates roughly equal to natural population growth, indicating
that employment creation for their hinterland labor force has been close to nil in most cases
(Mathur 1982; Kammeier 1984).

Two studies on the northern city Chiang Mai, then Thailand’s second largest city, were designed
to directly assess specific rural-urban linkages and flows. The first, conducted in the mid-1970s,
found that almost all of the economic impacts, including purchase of inputs for Chiang Mai
processing and production, demand for services in the hinterland, and increased local incomes
and expenditures, were reduced to zero or near zero levels beyond a one hour travel time
distance from the city (Bunchorntavakul 1976). The major reason for villagers to visit Chiang
Mai was shopping, but the frequency of visits even within a 90 minutes bus travel range from the
perimeter from the center of Chiang Mai) averaged only about one trip per week, and more than
one-third of the households said that they very seldom or never went to Chiang Mai.

A decade later a follow-up study of Chiang Mai was carried out using the same methodology
pioneered in the earlier research (Ladavalya and Siripool 1988). Carried out several years after
the government had explicitly designated Chiang Mai as one of only a handful of city's to receive
top priority planning and investment as a regional growth center, the study nonetheless found
that the spatial spread of economic and social benefits of the city’s growth rapid growth was still
very restricted. In fact, the expenditures by Chiang Mai factories accrued greater benefits to
Bangkok than to the hinterland of Chiang Mai. The study concluded that “the countryside
cannot provide what is needed in a modern, industrializing economy,” but it also noted that
“perhaps it could be said that no concerted effort has been made to develop new business that
can use what is available, such as agricultural products” (p. 54). It concluded that villagers
residing more than 1 hour from Chiang Mai “find practically everything they need in their own
villages and immediate areas, with only a few of their total expenditures going to Chiang Mai (p.
46), and noted the prevalence of itinerant merchants moving by pick-up truck from village to
village to sell basic commodities and even luxury goods: “the people can buy food, clothes,
medicine, kitchen appliances, candies” and other goods from the mobile vendors.

Thus although in comparison with the earlier study, greater proportions of nearby rural residents
were interacting with Chiang Mai, economic multiplier effects were still limited. Related
research on growth centers in five Asian countries -- Surabaya (Indonesia), Jaffna and Kandy
(Sri Lanka), Mariveles (Philippines), and Taxila-Wah (Pakistan) -- all found that diffusion
processes and multiplier benefits to rural hinterlands were not significant (Kammeier and Swan

These findings show that while planning for a single regional center over an extended period of
time may have pay-offs in terms of generating non-agricultural employment opportunities for
rural populations and increased local rural-urban interaction for social and consumer shopping
purposes, backward and forward linkages between town and countryside continue to be very
spatially limited. Although a focus on a very limited number of cities might save public
expenditures on infrastructure and service provisions, the Chiang Mai study suggests that it may
not be the most effective means of stimulating the development of rural regions. As a study in
the Southern Region of Thailand concluded, the “optimistic view that natural forces would create
a growth effect from Surat Thani City to its area of influence is, most probably, a dangerous
supposition” (Shiann-Far 1986). This calls into question the strategy of building the urban
system in a stepwise down-the-urban-hierarchy manner that, in effect, focuses on such a small
number of cities that vast rural areas are left beyond the range of benefits of urban growth. Such
a strategy misses numerous opportunities for developing a more dispersed pattern of regional
development via a number of smaller-scale rural-urban linkages and regional networks. Finding
alternatives to this type of growth pole approach needs to be given greater consideration if towns
are to play the development roles assigned to them.

2c Rural perspective on the role of cities

Most analyses of growth centers in rural development assume the perspective of the city looking
outward to its hinterland. Studies from the opposite vantage point in the village reveal quite
different insights. A study in Roi-et Province, Thailand, which looked at circular migration as
part of rural household economic strategy formulation, found that much of the interaction of
rural households with cities is temporary, often a month or two and not simply on an agricultural

slack season basis (Lightfoot, et al. 1983). Moreover, urban interaction does not
characteristically include the poorest of the poor, who are found to be without the necessary
resources to be able to move extensively over space. Rural-urban linkages are thus highly
selective in terms of access to urban functions as well as timing and duration of stays in urban

The implications of such research have been overlooked in most urban policy formulations.
Temporary or circular migrants do not tend to show up on annual household registration data and
are even less likely to be counted in decennial census-taking. Yet studies on this phenomenon in
Asia clearly show that temporary and circular migration is much greater than permanent rural-
urban migration. Furthermore, when seen from the household perspective, the view of the urban
world is not that of a single urban center, but is instead one of a network of rural and urban
linkages and employment possibilities, including destinations abroad. Rather than specific
destinations themselves, it is this network that provides the spatial frame for the formulation and
acting out of household economic strategies.

A popular argument put forth in favor of concentrating public investments in only a very limited
number of centers is the idea that the principal means of reducing rural poverty is not to focus on
“place prosperity” by trying to develop every region and city; the focus should be on
encouraging “people prosperity” through migration out of lagging rural areas to selected growth
centers of high potential. While it might be valid to say that the poor should not be artificially
trapped or prevented from migrating from low to high opportunity areas, caution needs to be
taken in advancing this view. Rural-urban migrants tend to be young, physically fit and better
educated than their village counterparts; the elderly and the poorest people do not leave “poor”

In this regard, the study in the Northeast province of Roi-et cited above found that migration to
Bangkok tended, over time, to erode village-based networks as migrants became "captured" by
urban-based obligations and remittances to home villages decreased. Villages also experienced
losses of scarce labor in peak agricultural seasons. Finally, remittances were not channeled into
community infrastructure and thus could not compensate for the decline in labor available for
community projects. Thus while some households gained from migration to large metropolitan
regions, short-term gains in consumption levels for individuals and their households might be at
the expense of the longer-term potential of the local economy (Lightfoot, et al. 1983:41). What
is needed, therefore, is a regional development strategy that captures these potentials to their
fullest extent. Such a strategy would have regional cities as one component, but would also gain
from including a more decentralized approach to rural-urban development at the local level.

2d Rural-urban interdependencies in regional planning

In view of the limitations of growth centers and urban functions in rural development approaches
discussed above, rethinking the role of cities in rural regional development raises the question of
how to bring rural and urban development potentials and complementarities together in the
planning process. A point of departure for addressing this question is to recognize that the
functions and roles played by cities in most rural areas are outcomes of interdependencies that
have no one-way urban-to-rural causality. Rather, rural and urban relations need to be seen as

being mutually reinforcing. These relationships are summarized in Table 1, which shows that for
every role of a city, there is a necessary role to be played by its hinterland.

As the table indicates, towns in rural regions act as higher-level market centers of agricultural
and rural commodities for both regional and extra-regional sales and distribution. Since the
town-centered marketing functions cannot exist without significant levels of marketable
surpluses being produced in rural areas, it follows that town and countryside are mutually
dependent. To expand production rural producers need marketing networks provided by towns
and the urban system; but without continued expansion of agriculture and agro-based processing
activities, rural towns cannot be expected to grow. Similarly, intensification of agriculture will
necessitate the appearance of shops in towns to supply increasingly sophisticated inputs and
repair facilities that a single village cannot economically sustain.

Continuing down the list of relationships in Table 1, a major source of growth of rural towns is
increasing demand for non-agricultural commodities for rural household consumption. Research
has identified this as the single most important factor for the growth of rural towns (Gibb, 1986;
Somluckrat, 1990). As in the other relations, the ability of towns to act as consumer convenience
centers rests on increasing rural prosperity and rising real incomes not just for a few farmers, but
for the majority of rural households.

Table 1 Urban and Rural Linkages and Interdependencies

URBAN <-----------> RURAL

• agric. trade/transport center <-----------> • agricultural production

• agric. support services <-----------> • agricultural intensification
-- production inputs -- rural infrastructure
-- repair services -- production incentives
-- information on production -- education and capacity
methods (innovation) to adopt/adapt innovation
• non-agric. consumer markets <-----------> • rural income and demand
-- processed agric. products for non-agric. goods and
-- private services services
-- public services (health,
education, admin.)
• agro-based industry <-----------> • cash crop production and
agricultural diversification
• non-agric. employment <-----------> • all of the above

2e Targeting the impacts of urban development

As the many studies reveal, investing in cities to promote the development of rural hinterlands
encounters so many intervening factors that actual beneficial impacts are often much less than

expected and unevenly distributed among rural households. Banks may not lend to high-risk
rural producers; schools may require informal fees that put them out of reach of poorer
households; roads may accelerate environmentally dangerous resource extraction.

Given the likelihood that the realities of rural-urban relations may deflect efforts away from
achieving their goals, there is thus a manifest need to more clearly specify who is to benefit from
urban investments and what are the means to ensure the desired outcomes. This would not only
require more thorough study of rural development potentials; it would also call for closer
attention to monitoring programs to ensure that target groups, such as small farmers, landless
workers, and the rural poor are actually reached.

3. Regional Networks/Clusters as a Spatial Framework for Rural Development

The purpose of the foregoing discussion has been to demonstrate that prevailing theories and
policy models do not capture either the realities of or the potential for rural regional
development.. There is thus a need for a new paradigm of spatial development for policy
formulation. Such a paradigm would have to overcome a number of major obstacles, including
the dichotomization of planning into rural and urban planning bureaus that promote rivalry rather
than collaboration and administrative divisions that separate cities from their hinterlands in
planning and management at local levels. In addition, mechanistic models of spatial and
development processes, most of which focus on urban nodes rather than rural regions, need to be
put aside in favor of efforts to include local variations in rural-urban linkages in identifying
components of a national spatial system.

By way of exploring how a new paradigm of rural regional development can be constructed from
local level research on rural-urban linkages, Table 2 and Figure 1 contrast the growth pole
concept with an alternative regional network (cluster) concept, the latter of which incorporates
rural and village structures with rural-urban linkages and flows. Table 2 shows, first, that
whereas growth poles have been single-mindedly focused on urban-based manufacturing as the
leading sector for regional development, a regional network approach recognizes the multi-
sectoral nature of local level development in rural region and acknowledges the role of regional
resource endowments and already existing activities rather than limiting the prospects for local
development to inducements to decentralize footloose industries from core regions. Building
networks allows for a variety of sources of economic growth and does not assume that each will
be urban based. Bulk-losing processing and agro-industry, for example, may be more efficiently
located near the fields or along major transport routes, including waterways, rather than in cities
or towns.

Secondly, most growth pole policies adopt an implicit top-down hierarchical view of the world
and, in so doing, tacitly assume that city-size can be used as the principal indicator for
identifying a region’s principal city and, further, that cities of the same size will have the same
functions. In contrast, the network concept draws from research in Asia and elsewhere showing
that not only is city size an inadequate indicator of either growth potential or local linkages, but
also that cities of the same size class can have very different functions and development profiles
(Cohen et al. 1980). Rather than trying to make a single, large city into an omnibus center for a
vast region, the network concept is based on a clustering of many settlements, each with its own
specializations and localized hinterland relationships. Thus small towns in upland areas or a

region may be key marketing centers for estate crops; another may be near an important cultural
center; and yet another may serve as the administrative center for the region.

Third, growth pole approaches tend to view rural areas as backward and dependent upon
diffusion or trickle-down impulses from cities for their development. The network approach
takes a more even view by recognizing that appropriate investments in agriculture can lead to
higher levels of per capita income in that sector and, further, that rural prosperity is a powerful
source of urban growth in agrarian regions. At the same time, selected investments in region-
serving functions in cities can help to raise the economic growth potential of both town and


1. Basic Urban-based manufacturing; usually all sectors, depending on local
Sector focuses on large-scale ‘propulsive’ regional endowments and conditions;
industries and ‘footloose’ production emphasis on local small-medium size
units headquartered outside the region. regionally-based enterprises.
2. Urban Hierarchical, centered on a single Horizontal, composed of a number of
system dominant center, usually identified by centers and their hinterlands, each
population size and associated with the with own specializations & comparative
assumptions of central place theory. advantages.
3. Rural- Image of diffusion processes moving Image of a complex rural-urban field of
urban down the urban hierarchy and outward activities, with growth stimuli emanating
Relations from the city/town to its rural periphery. from both rural and urban areas and with
Rural areas as passive beneficiaries of the intensity increasing along regional
“trickle-down” from urban growth. inter-settlement transportation corridors
4 Planning Usually top-down via sectoral planning Implies the need for decentralized
style agencies and their field offices. Regions planning systems, with integration and
have “misty” boundaries determined by coordination of multi-sectoral and rural
economic interaction. and urban activities at the local level.
5. Major Industrial decentralization incentives: agricultural diversification, agro-industry,
policy tax holidays, industrial estates, national resource-based manufacturing, urban
areas transportation trunk roads. services, manpower training, local
inter-settlement transportation networks.

Fourth, a major difference between growth pole and regional network concepts is found in the
styles of planning they encourage. In the former, the urban node is the all-important spatial
“actor.” Demarcating its hinterland is presumed to be a fruitless and unnecessary task.
Regional boundaries remain “misty.” In contrast, the multi-sectoral nature of the cluster concept
implies the need for a localized capacity to coordinate a large number of interrelated activities.
Using existing provincial or district level boundaries as the level at which coordination and
integration of planning is to take place thus becomes an important aspect of the cluster approach.

Finally, the two concepts also differ in terms of the types of policies they are likely to consider.
Growth pole policies are built around incentives to attract manufacturing, namely, the public
provision of economic infrastructure for urban industry and the expansion of national trunk road
expansion linking growth centers with the capital city and, by extension, the international
market. In contrast, the regional network concept recognizes the need to expand rural as well as
urban infrastructure, and also gives stronger emphasis on local road and transportation linkages

among major centers, villages and towns within the region. Equal attention is given to social as
well as economic overhead capital. The provisioning of piped water, electricity, sewage,
drainage, schools, and improved health services is not merely a matter of ‘welfare’ but is also
important to the upgrading of the quality of regional life needed to sustain economic growth as

The contrasts between the growth pole and regional cluster paradigms are schematically shown
in Figure 1. The network concept is based on three principal considerations. First, variations in
rural-urban linkages are great even among in the hinterland of the same principal town; a
clustering villages and towns into a regional unit of development can therefore take advantage of
the diversity as well as complementarities among various centers and between each center and its
immediate hinterland within a given region. It does not rely on a single center to lead regional
growth. Relations among centers are more horizontal, complementary and reciprocal.

Second, such clusters already exist, even if in a rudimentary form. Interaction among villages
and towns, rather than being confined to dyadic relationships between village-town pairs, form
localized networks with varying degrees of intensity across the region and beyond. Such clusters
can be identified and demarcated in the initial instance by using existing flows of goods and
people among settlements. In the case of Indonesia, for example, regional networks were
identified by using data showing the frequency of bus, truck and automobile movements among
settlements (Douglass 1984).

The third consideration is that a cluster of well connected and highly interactive rural and urban
settlements may be better able than a single growth pole to provide a level of agglomeration and
economic diversity to act as an antipode to the growth of core metropolitan regions. Given the
reduction of time distances among settlements made possible by modern transportation and
communications systems, somewhat dispersed towns and villages can be linked together to form
an effective range of daily interaction that would have been impossible in most rural regions only
a decade or two ago. A more robust regional economy covering a wider spatial scale than a
single town and its hinterland and offering a wider array of economic activities can also more
readily weather the vagaries of external price shocks and shifts in demand and resources.
Greater potential would also exist to capture upstream and downstream linkages and multiplier
effects in the region.

In sum, the regional network approach advocates a more decentralized system of planning for
working out the dynamics of rural-urban linkages. It challenges many of the precepts of the
center-down diffusion and urban systems models that have dominated policy formulation. It
suggests a more diversified, multi-stranded approach to regional development which rests on
integrating rural with urban development at the local scale. The approach being recommended
advocates a disaggregated and finely-tuned set of policy interventions that allow for variations in
regional resource endowments, existing divisions of labor in urban and rural sectors, and local
development needs and potentials.

In Indonesia, the transition from a rural to an urban-based economy is now well underway and is
reaching into all regions in the country. The timing and pace of this transition is, however,
highly uneven over space. If the objective of ameliorating the economic consequences of such
uneven patterns of development, which favor very large urban agglomerations and particularly

the greater Jakarta metropolitan region (Jabotabek), is to be pursued, the general task at hand is
to find ways in which the dynamics of rural-urban linkages and transformations can unfold in a
more complementary and reciprocally beneficial manner over the national territory. The
regional network model is proposed as a framework for this task.


Growth Pole Regional Networks

Industrial Diffusion Model Rural-Urban Interdependency Model

Core Region

National Extended
Regional Regions

-based Regions

Rural Towns and Villages

Regional Clusters as Complex Rural-Urban (Proto-) Agglomerations

agro-processing center industrial estate
administrative center tourism center
agric. input/consumer ag. market
convenience center
rice/grain resource mining
fruit growing animal husbandry

4. Regional Variations in Rural-Urban Linkages: the Case of Indonesia

The regional network paradigm allows for a much wider diversity in rural-urban linkages at the
local level than do prevailing models. Evidence from visits to rural areas in Indonesia show how
urban-regional systems have much less regularity and uniformity than either central place or
industrial diffusion models allow. Historical factors related to the development of each city and
region, the presence of government, and the nature of linkages with the national and international
economy add to the diversity of possibilities.

4a Dimensions of regional variations
Figure 2 identifies key dimensions of regional variations in rural-urban linkages. On one side are
those elements in the realm of society and economy. Socio-economic relations, for example,
differ in terms of relative and absolute levels of inequality in income, access to land and other
resources, and types and levels of skill attainment by the regional population. These differences
may be expressed in terms of class divisions, ethnicity, gender or other socially created
cleavages. Unless explicit attention is given to such differences, improving the role of the city
may widen rather than close gaps. In Bangladesh, for example, the widely-acclaimed Grameen
Bank was established principally to reach poor rural women who were without sufficient
collateral or social status to qualify for credit from either banks.

Urban functions also vary according to the structure of the rural economy. Fishing villages
might require cold storage centers; whereas cattle raising areas might need livestock holding and
feeding lots at key transshipment centers. The economic structure of a town reflects the capacity
of the region to capture upstream and downstream linkages and multiplier effects. Thus a
regional economy based only extractive activities with no local processing creates an urban
profile markedly different from one of a town in a more complex region.

The organization of production (production regimes) within a given sector or activity will also
have a substantial bearing on rural-urban linkages and the requirements for urban services and
infrastructure. Plantation versus smallholder cashcrop production, for example, show
significantly different town profiles. Plantations, which typically use full-time low-wage labor
and ship produce in large bulk and may provide lodgings and company stores for workers, may
show little demand for local town services; in contrast, a smallholder production regime may
depend to a greater extent on towns for supplies, markets, cooperatives, and consumer shopping.

Variations in rural-urban linkages also fall into a second realm covering the habitat and human
settlement system of a region. Conditions of the natural environment, such as deforestation
leading to constant flooding, may threaten the existence of the town itself, and beyond that it will
influence patterns of access and, therefore, the types of functions supported by the hinterland.
The natural resource base is perhaps the single most important determinant of a rural regional
economy and, therefore, the role of towns in that economy. The classic case of single-resource
regions in which towns experience a “boom-bust” life cycle is an extreme example of this

Conditions of the built environment are major sources of regional differentiation in rural-urban
linkages. Regions with high grade irrigation allowing three rice crops per year have much higher
levels of prosperity and commercialization than do most rainfed, single crop regions. Local road
networks also profoundly affect rural-urban flows of people, goods and services. Electricity,
telephone services, and other linkages add to the physical development of markets and town and
village infrastructure to transform the nature, scale and magnitude of rural-urban interaction.
Finally, local rural-urban linkages also vary according to relationships to the larger spatial
network of towns, cities, transportation and communications flows. As noted, many studies
show, for example, that local towns may be bypassed as centers for agro-processing, markets and
other urban functions as transportation linkages to more distant, larger towns are improved.


Economic Policies & Political Institutions


Socio-economic Relations Settlements & Habitat

Socio-cultural Natural
and Economic Environment
Relations & Resource Base

Local Variations in Spatial

Economic Rural-Urban System &
Structure Linkages Linkages

Production Built
Regimes Environment


• Basic needs poverty levels Environment
• Inequality in distribution of income and assets • Composition and diversity of flora and fauna
• Level of landlessness, marginal farmers, tenancy • Ecological integrity
• Basic skills by income groups and gender • Vulnerability to natural disasters
• Access to basic social services by income class Resource base
STRUCTURE OF RURAL ECONOMY • Soil quality and degree of erosion
Composition of basic sectors (GRDP) • Water availability
• Primary: mining, fishing, forestry • Forest reserves/minerals
• Secondary: agro/resource-processing; manufacturing BUILT ENVIRONMENT
• Tertiary: tourism, commerce, services • Farm (irrigation, drainage, storage)
Upstream-downstream multipliers • Village (roads, bridges, electricity, water supply,
• Local production of inputs communications, housing)
• Processing/Agro-industry • Urban (village-town roads, market centers,
• Commercial sales; producer services, consumer services water, sewerage, electricity, communications)
Distribution of labor force • Regional (trunk roads, elec., communic., market centers)
• Participation rates and dependency ratios SPATIAL SYSTEM/LINKAGES
• Distribution by sector, occupation, status • Rural population density
RURAL PRODUCTION REGIMES • Levels of urbanization and complexity of urban system
Land Distribution & Property Regimes (no. towns/cities, traffic flows, air and water transport)
• Smallholder vs. large (capitalist/socialist) farms • inter-village & village-town access (road quality & frequency
• Agribusiness: plantations, contract farming of public transport)
• Common/state/private land distribution • communications linkages
Producer Organizations (e.g., cooperatives) • circular migration in the region and with other regions

Since this is a dynamic process, so too is the changing functional profile of a town. Activities
that were viable at one point in time die out as the larger spatial system develops. More than just
a question of access to markets, the connectivity of settlements presents a matrix of competition
among regions in terms of comparative advantage and economic potential. Linking up more
isolated regions, which are normally characterized by high levels of economic self-sufficiency in
their hinterlands, with national and international networks would, for example, be expected to
profoundly change the nature and role of towns in the region.

4b Rural-urban linkages in the contemporary Indonesian context

In giving explicit attention to rural-urban linkages, the Banjarnegera study mentioned previously
is perhaps the only one of its kind in Indonesia. Even this study did not, however, cover the
range of regional variables summarized in Section 4a above. Much research has nonetheless
been carried out on some aspects of rural-urban linkages in regional development. Among the
more useful studies in this regard is the longitudinal research carried out in eight villages in Java
over a 25 year period from the early 1970s. Although not concerned with cities per se, the
research contains a wealth of information about rural transformations occurring on Java during a
period of intensive urbanization and quantum expansion of rural-urban transportation networks
and rural-urban interaction. In one of its most recent reports (Collier 1993), the study found
significant differences among villages were the result of not only land capability and other
aspects of village potential, but were equally patterned by relative distance from major
metropolitan centers. Three types of rural regional contexts were discovered:

(1) Villages within 60 km of large rapidly industrializing cities were characterized by

high daily commuting (more women than men, which is reflective of the transnational
investment in female-dominated export-processing industries driving metropolitan
economic growth in Indonesia) to factory and other work on the urban fringe. Much
of lowland Java falls into this category with good roads to most the villages and
trucks and buses routinely entering them every hour or more frequently (even every
minute). In many instances factories sent buses to pick up workers every day.
Communications had become internationalized, with TVs accessible to many village
households. Crop regimes had changed in response to increasing labor shortages at
peak periods, and wage rates for agricultural laborers had increased. Agricultural
land prices were also rapidly increasing in these areas, with land sales increasing to
non-village buyers.

(2) Villages in major rice producing areas perhaps near a city but beyond the 60 km.
range commuting range of those rapidly industrializing showed monthly and yearly
migration for urban employment and had more full time farmers operating small plots
of land. Agricultural labor shortages existed and were caused by yearly migration.
Large farmers were very commercially minded. Wage rates increased in villages
having alternative job opportunities; otherwise they were found to have stayed
constant. The quality of village housing was improving, but a substantial number still
had dirt floors and bamboo walls. Unemployment was found to be low (mostly
among more educated males who refused work in agriculture). Most children were
finishing at least junior secondary (SMP) education; and slower population growth
was leading to a decline in school-age children at lower levels. Electricity had

become widely available, but not to all and not to more remote villages; some had
TVs; daily newspapers were reaching villages; and at least daily truck and bus
services were reaching most villages. Compared to a decade earlier there were major
increases in standards of living, although other reports show this to be uneven
(Douglass 1997).

(3) Villages in upland areas distant from large cities were found to still have poor
communications and continued to be dependent on agriculture because of distance to
off-farm employment opportunities. The standard of living had fallen in these villages
due to dependence on a few crops with declining prices (clove and coffee). Younger
and better educated people were migrating out to work as laborers in factories and
construction and as agricultural laborers in lowland areas in rice and sugar cane. Out-
migration was leading to a switch to less labor-intensive (tree) crops. Electricity was
being extended, but not to the extent of other areas. Village industries were
increasing (roof tile, brick making, clove leaf refining, handicrafts; agro-processing
(red cane sugar), but mostly in low-productivity, low-income activities. Televisions
had begun to appear, but were not yet widespread.

In light of these findings and the scarcity of other research on rural-urban linkages, visits were
made to several of villages that have been identified under the special Presidential Decree for
rural poverty alleviation (IDT). Unlike other poverty programs, the IDT program has two key
features. First, it is not aimed at poverty areas as measured by income or basic needs, but rather
as having inadequate levels of infrastructure and connectivity with towns and regional
transportation networks. Secondly, it is intended to be highly decentralized, with districts
(kecamaten) being the level at which projects are to be identified. In September 1996 three
candidate IDT villages were visited in the outskirts of Yogyakarta in Central Java; one was
visited in eastern Lombok, and another was visited in the hill areas near Bima on the island of

4c three villages near Yogyakarta

Two of the villages near the city of Yogyakarta were actually quite well connected by roads and
transportation, had electricity, and were relatively prosperous. The third, located within a 30
minute drive from Yogyakarta and in the hills of the historic Prambanan area of Hindu temples
and palaces, had very poor road connections and much lower levels of income and welfare.

The city of Yogyakarta has itself experienced a boost in prosperity in recent years, but not so
much from its agricultural hinterland as from its role as a national center for all levels of
education. Schools and universities are still being built in the city, and parents are sending
children from across the country to study in this relatively safe city with a high reputation for
excellence in education. As a result, the economies of many villages surrounding the city seem
to be increasingly specializing in products for the urban population of Yogyakarta as well as for
export to other cities. The economy of one nearby village, for example, is dominated by the
growing of salak pondok, a fruit that can be grown widely throughout Java, but is reputed to
have an especially good taste when grown in this village. The growers are not all village people,
but also include urbanites who hire villagers to take care of daily management such as guarding

their crop from thieves. The ribbon of development along the principal road through the village
is dominated by roadside stands selling the fruit, as is the major village market.

Not very far from this village is another IDT designated village that specializes in pottery. The
owner of one of the principal shops told of how in the 1970s the lack of a local market for his
wares in the region led him to load his pottery on his bicycle, had it put on a truck and
transported to big cities such as Surabaya where he would cycle to sell his products. In the
1980s he decided to stop selling in other cities and to open a shop in his village. He is proud that
he now exports not only to other destinations in Indonesia, but even to cities in Australia. Other
pottery shops have prospered in the village as well, which is now well known as a pottery
village. Somewhat surprisingly, the village has no special clay resources that distinguish it from
others in the region. It seems that the entrepreneurial efforts of village residents combined with
the growing prosperity of urban residents in the education-driven economy of Yogyakarta have
benefited from improvements in connectivity with cities in Central Java to create the market
niche for this village.

The third village, located in a dry hilly area without paved roads, was also specialized in a single
export, which in this case was limestone slabs quarried from land underneath rice fields and used
in urban building construction. A new road had been connected to the village, but it was not yet
paved. Nonetheless, it now allowed large trucks to come to pick up the stone. Until that time,
each of the large, heavy rectangular stones were carried down the mountain on the heads of
village people. Although the trucking may have reduced some of the work, it is not clear that it
has resulted in higher incomes. Rough calculations suggest that at the end of the day the
villagers earn no more than the going wage for landless laborers in the region even though they
own the quarry land. Since they do not own the trucks, there is a likelihood that they face
monopsonistic pricing of their cut limestone by the buyers from the city. This pattern of
connecting up with larger cities but gaining little economic benefit due to monopolistic and
monopsonistic markets is even more pronounced in Eastern Indonesia away from the densely-
settled rural regions of Java.

Crop production in the upland village near Yogyakarta was severely limited by water
availability. Only one crop of rainfed rice could be produced per year, and this was insufficient
for household needs; the village was thus a net buyer of rice and other food crops. Limestone
sales were the only source of income for these purchases. Given the poor agricultural conditions,
IDT funds were used by households to buy calves which are to be raised (but not used for
farming) and sold for a profit outside of the village. Households were being equipped with their
own cement water storage tanks through government assistance, but most were almost dry in
September. Houses were constructed of poorer materials, and consumer durables were limited.

What is somewhat surprising is that although this village is less than 30 minutes from
Yogyakarta by automobile, its level of interaction remains low in many important ways.
Households reported, for example, that they almost never went to Yogyakarta for any reason.
Although work might be available there, they said that for historic reasons they had never
developed social relations or networks in the city and thought that without such relations no one
would give them a job. This sentiment is consistent with most migration studies, which find that
job search through migration follows rather narrow social pathways and that specific job
opportunities at the destination are usually known and even agreed upon before migration occurs.

In many cities, specific jobs such as pedicab driving or ice cream vending are often associated
with migrants from a single region or village.

Children in this village had to walk to school, and due to the lack of water during the long dry
season, cattle had to be walked down the mountain for grazing and water. Should the roads be
improved and daily transportation be available between this village and Yogyakarta, it is not
clear that the results will be a widespread improvement in the village economy. Lack of social
contacts in the city, the controlled piece rate of the stone work, and lack of other resources in the
village coupled with the lack of water do not suggest that roads will be enough to generate a vital
economy. Perhaps if the stone cutters owned their own truck or were able to form a cooperative
they might have more leverage over buyers. In any case, direct attention to social and economic
factors would need to accompany infrastructure planning if the village is to genuinely benefit
from improved rural-urban transport linkages.

4d “Remote” villages on Lombok and Sumbawa

Lombok and Sumbawa are included in the larger region of Eastern Indonesia, which is
considered to cover the most isolated and poorest localities in the island nation. Lombok and
Sumbawa are relatively densely-settled compared to other major island groups, and in some
locations the densities approach those of Java. On both islands, the variations are very great in
terms of resource base, economic potential and linkages to the rest of the national and
international economy. On Lombok, the most profound economic change now underway is the
rise of international tourism centered on the Senggigi beach area, which currently accounts for at
least 40 percent of the economy of West Lombok. Tourism is, however, limited mostly to one or
two peak seasons (August and the end-December to early-January period) and is low throughout
most of the remainder of the year. Even so, with the many white sand beaches and the proximity
to Bali, tourism is likely to be the basic and leading sector of growth for the island for years to

For the rest of the island, tobacco is the most widely grown cash crop. Although technical
irrigation exists, many of the more arid parts of the island still depend on rainfed irrigation for
paddy and other crop growing. Almost none of the tobacco is dried or processed either at the
village level or even on Lombok, but is instead shipped directly to Surabaya (Java) for
processing and further marketing. As such, very little of the total value added of tobacco is
captured on Lombok, leaving the potential for high levels of reinvestment and economic
diversification limited. Farmers say they would double their income from tobacco if they were
allowed to dry it before selling, but apparent collusion among buyers prevents this.

One of the most significant features or the economy of many Lombok villages is the emigration
of workers to Malaysia, which involves professional recruiters and, at least at the present, is
allowed by the Malaysian government. In some villages, these migrants account for almost one-
fifth of the total labor force. Remittances from these migrants are reportedly important sources
of cash in the villages, not a few of which claim to have been depleted of adult men.

Two villages, Menies and Pengkelakmas, located in East Lombok have approximately 700 and
1200 households, respectively. A bridge and secondary road connecting the villages to the
kabupaten (regency) road were to have been completed by the middle of 1997, which will for the

first time enable trucks to come to Pengkelakmas to greatly improve flows of interaction with the
village, including the transport of its agricultural commodities. Currently, it takes one day for
villagers to carry their products, such as tobacco, to the main road for delivery to middlemen.
Fisherman in the adjacent village also hope to be able to sell their catch to urban centers on the
island. With increased market potential, incomes are expected to go up as well as more products
can be sold in a shorter space of time. Since even a large increase in the scale of production
from the village would not be expected to affect the market price, selling more is not expected to
result in lower prices unless certain monopolistic practices among buyers prevail to push prices

The villages of Samboir Lama and Teta, located in the hills above Bima in Nusa Tenggara Barat,
have just had new roads built that now links them to the kabupaten road. As with all villages
receiving road projects, expectations are high that commercial sales of village products (rice, red
onions and garlic) will go up and, along with increased access to schools and public services, so
will general income and welfare. Minibuses now come three times a day to the villages,
although they do not as yet seem to be well timed for the two-shifts of school, which takes
children about 3 hours walking one way to reach. The minibus takes only approximately 20
minutes to reach the same lowland vicinity of the school.

The city of Bima itself provides an interesting combination of factors that account for its relative
prosperity. As the center of a Sultanate, it appears to enjoy a higher level of political, and
therefore public spending, attention than do many other towns of its size. The city is well laid
out and appointed with a high level of road and other infrastructure. It is also the principal point
of connection and stopover for tourists going to Komodo Island, home of the huge Komodo
Dragon lizards. There are as yet very few hotels, however, but new ones are being constructed
and with a more reliable air connection to Mataram and Denpasar, the international tourist
economy is expected to enjoy steady growth. One of the hotel managers said that programs have
been organized to recruit and train talented village youth for the hotel industry. Shops in the
town include not only numerous clothing and dry goods stores, but also durable goods stores
selling televisions, stereos, electronic appliances at a scale that suggests that the town has a
sizable middle class, most of whom come from the more than 2,000 thousand public employees
in the town, which includes teachers and health workers as well as government planners and

In sum, although the region exports onions and beef to Java, there is little or no processing of
these products in the region, and the vitality of the town appears to depend as much -- if not more
-- on the large size of the civil service and government administration. As with Yogyakarta’s
education economy, this can, however, have a somewhat catalytic effect in raising demand for
staples and higher value added food crops from its rural hinterland than would otherwise be the
case if the prosperity of the town were wholly dependent on rural demand for its services.

4d Summary of observations
Drawing from available case studies in Indonesia and comparing the three villages around
Yogyakarta and those on Lombok and Sumbawa, a number of points are revealed that show how
regional differentiation occurs at the micro-spatial (village) as well as at the macro-spatial
(region and above) scale:

(1) Specialization in markets occurs at the village scale, even in poor villages and even
among villages in the same locale.

(2) Rural-urban interaction can be very low even with villages located very close to very
large cities. This low interaction is the result of poor transportation linkages as well
as specific socio-economic conditions of a given village.

(3) Improving spatial linkages through, for example, road development, has an
indeterminate impact on villages. While access to schools and other services may be
heightened, positive income impacts are less certain for several reasons:
monopolistic practices may still prevail in buying village products; increases in
supply may dampen per unit prices; the village may lack comparative advantage in
marketing its products as other villages also begin to enjoy greater access; resources
may be depleted in an environmentally unsustainable manner.

(4) Village potential is imbedded in both “absolute space” -- the nature of its own human
and physical resources -- and “relative space” -- its degree connectivity to larger
urban and regional systems. It cannot be solely implied by either natural resource
endowments, although these are important, or by past economic histories. The
success of the pottery village near Yogyakarta, for example, would not have been
predicted on these bases.

(5) More successful villages are linked to and interact with more than one urban center.
Access to regional and national urban networks widens market potential and may also
help to overcome monopsonistic markets. It also provides a wider range of
alternative income-earning opportunities for various members of rural households.

(6) Rural-urban linkages are quite specific and do not generally include a full range of
activities. In the case of the upland village near Yogyakarta, for example, the major
linkage was limestone trade, with limestone being sent out of the village several times
per day by truck. In contrast, flows of people and goods between the village and
Yogyakarta were highly restricted, as was information about job opportunities.

(7) Information about linkages may be more powerful from distant rather than from
nearby or even national sources. Recruitment of labor from Lombok for Malaysia did
not rely on information coming from cities in Indonesia, but came instead directly
through migrant networks from Malaysia to specific villages.

(8) Different types of flows between village and town will therefore develop at variant
rates with somewhat distinct spatial patterns. The movement of people is unlikely to
follow the same paths as the movement of goods or finance. There is, however, a
general spatial matrix of interaction within which many of the flows are contained,
and within this matrix cities and towns begin to play their own roles. As this matrix
expands and changes through local, national and global impulses, so will the role of
rural towns.

(9) The growth and prosperity of key rural towns, particularly ibukota (municipalities)
and administrative centers off Java, appears to heavily rely on the size and expansion

of the civil service. As such, government sector employment and spending should be
included as an important factor in rural-urban linkages. These unexpected sources of
local demand can also heighten the potential for rural production of higher value
added agricultural products, processed food, and village crafts.

(10) The emergence of local networks of villages and towns is most easily observed in
traffic movements. Two major factors explain these networks. First, as mentioned,
crop differentiation at micro-levels stimulates local as well as longer-distance
exchanges among settlements. Second, public village-town bus systems provide a
critical level of daily interaction needed to bring village households into frequent
contact with towns and other villages. Port and other types of transport development
add to this process.

(11) Rural regions off Java seem to have very low levels of backward and forward
linkages between village level production and the rest of the economy. Most
production appears to be sent to ports for shipment to Java or abroad without even
simple processing. National government regulations requiring processing to be done
on Java even when it could more efficiently (and with greater regional equity) be
done near the source, particularly in the case of bulk-losing processing or processing
of perishables which are typical of many rural products, is an often-stated reason for
the absence of local processing. The inability to localize these linkages is one of the
greatest limitations in promoting a more dynamic process of rural-urban development
off of Java. They also promote wider disparities as higher value added in processing
is captured in national core metropolitan regions.

(12) Financial flows to villages are typically on a people-to-people (e.g., migrant

remittances) rather than on an institutional basis (banks and credit associations). Rural
household access to low-cost credit is one of the weakest rural-urban linkages.

(13) Flows of information have experienced an extraordinary widening of scope through

the pervasive diffusion of televisions to even low-income villages in some regions
(notably across Java). While this has opened access to national and international
information about political, economic and consumer affairs and opportunities, there is
little local content, such as agricultural prices or other local producer information, to
enhance the knowledge and bargaining position of especially smaller rural producers.
Adding local information channels directed toward informing rural producers of vital
opportunities and market conditions would be a significant contribution to their
economic welfare.

(14) The environmental and ecological implications of extending rural-urban linkages

seem to be poorly understood or integrated into urban and regional planning. The
exploitation of natural resources from distant urban centers is responsible for a great
deal of environmental destruction throughout the world, and the expansion of cities
into rural peripheries is also cited as a major source of land degradation. Increasing
the potential for economic diversity in rural regions could, however, contribute to
efforts to move away from dependency on over-exploitation of the natural resource
base. In either case, more clearly understanding the environmental implications of

rural-urban interaction would make an important contribution to sustaining local

(15) In general there is a clear sense that rural-urban interaction, particularly that related to
road development and television, has experienced substantial increases everywhere in
the past two decades. While they have been instrumental in raising levels of welfare,
these linkages have not necessarily narrowed either rural-urban or regional disparities
in income. Accomplishing this goal would require the creation of virtuous cycles of
rural-urban development at the local regional scale (see Section 5 below).
Infrastructure, while of great importance, is not sufficient for this task by itself.

5. Toward an Agenda for Policy Research

Insights from real world contexts begin to sketch an agenda for policy research on rural-urban
linkages. Figures 3 and 4 attempt to synthesize the issues they raise by contrasting a virtuous
cycle of mutually reinforcing rural-urban linkages with a truncated form of interaction that leads
to lower levels of both urban and rural performance at the local level. In the virtuous cycle
(Figure 3), both (A) international and (B) national conditions are favorable to (C) regional
investment in (D) basic/leading sectors and the internalization of three key multiplier effects:

(1) employment generation directly through hiring in the export activities and indirectly
through hiring in related processing and input supplier firms;
(2) processing and manufacturing of basic products are carried out in the region; and
(3) local purchase of inputs.

Multiple sources of employment generate a widespread increase in household incomes, which in

turn generate increased demand for urban services and functions. The economy diversifies and
deepens through each cycle, with new leading sectors emerging as older ones run their course.
Local planning and producer responsibility combine to renew the environmental base of the
In contrast to a virtuous cycle of sustainable development, Figure 4 shows what may be the more
typical case in rural regions, particularly those located away from major metropolitan centers.
Because the region is overly specialized in one or a few primary exports, even somewhat minor
swings in international commodity prices have a volatile impact on the local economy. Regional
infrastructure and services also typically receive less priority than metropolitan regions and are
continuously below needs and potentials. The basic sector has few forward or backward
linkages, which greatly limits employment and household income increases. The poor
performance of the rural economy has a dampening effect on the local urban economy and its
growth potential. Dependence on a narrow resource base leads to over-exploitation and
environmental degradation. Low incomes and poverty persist. Towns are often more dependent
upon government administration and spending than they are on the rural economy. Even where
an enormously profitable mining or other extractive activity, such as forestry, exists, it does not
articulate well with its rural hinterlands, which in Indonesia may be composed of small-scale
societies that cannot be readily tapped to fill the new labor demands. They remain largely
enclave activities more linked to international rather than local markets. While cities around
these activities may boom, their impacts on rural areas remains problematic and register striking
negative and positive impacts.

While many regions may fall between the extreme examples in Figures 3 and 4, the diagrams
illustrate the ways in which rural-urban linkages can follow quite different trajectories. If one
purpose of national planning is to better insure a more reciprocal relationship between rural and
urban development, efforts need to be combined to foster a virtuous cycle of development.


A Favorable international commodity prices/
Localization and diversification of foreign investment
B Adequate provision of infrasturcture and basic services/
Support of local economic initiatives and organization



1 Primary and 2 Processing/ 3 Demand for

Non-primary Manufacturing Inputs

4 Increased
Household Incomes


5 Growth of 6 Growing Sales

Centers for Consumer of Inputs/Producer
Shopping Services

7 Rising Demand for Expanding
Health, Welfare and Marketing of Regional
Leisure Services 'Exports'

Renewal of Economic Broad-based
Resource Base/ Diversification/ Increases in
Environment/ Increasing Income and
Ecology Productivity Welfare

Research in a number of regional contexts in Indonesia is thus needed to better pinpoint
bottlenecks preventing opportunities for these efforts to bear fruit. Toward this end, Figure 5
presents a simplified template for research on rural-urban relations. By dividing the research
components into structures and flows, it suggests that rural structural change and development is
linked to urban functions and roles through a set of flows between rural and urban areas. The
task of research is to analyze (1) patterns of flows and (2) their combined impact on fostering
rural regional development (i.e., both town and countryside).

A Volatile/falling commodity prices/
Inputs imported, single resource/crop nvestment
B Inadequate infrasturcture and basic services/
Poor support of local initiatives and organization




1 limited, low- 2 No local down- 3 limited local

skill employ- stream linkages input demand
ment creation

4 Stagnant/falling
Household Incomes

5 6

Infrequent Limted Sales

Shopping in

of Inputs/Producer
Local Towns Services

7 Ineffective 8 Regional
Demand for 'Exports' bypass
Consumer Services Local Towns

Depletion of Single-resource/
Resource Base/ Monoculture/Pro-
Income Increases/
Environmental ductivity Gains
Persistence of
Degradation Realized

Five types of flows are identified in Figure 5: people, production, commodities, capital, and
information . Each has multiple components and impacts. Commodities, for example, take the
form of production inputs, marketed rural products, and daily wage goods (non-durables) and
durable goods for final consumption. Each may have quite different spatial linkage patterns as
well as variable benefits to rural and urban areas.




• labor commuting/migration FUNCTIONS/
• other migration (e.g., education)
• shopping/visiting/selling
• Non-agricultural
• Socio-economic PRODUCTION 2 employment
Structure/ • upstream linkages (inputs)
Relations • downstream linkages • Urban services
(processing, manufacturing)
• Rural Economy • Production supplies
• inputs • Non-durable and
• Rural Production • consumer non-durables/durables durable goods
Regimes • rural products
• Markets for selling
• Natural CAPITAL/INCOME rural products
Environment • value added
& Resources • savings/credit • Processing/
• migrant remittances manufacturing
• Infrastructure
Built • Information on
Environment INFORMATION 5 employment,
• production/sales/prices production, prices,
• welfare/social/political welfare services
• employment

• Agrarian reform POLICY INTERVENTIONS • market centers

• Agriculture intensification/ • commercial outlets
diversification • Roads/transportation • urban services
• Cooperatives • Electricity • banking/credit
• Enviromental programs • Communications • urban infrastructure
• Irrigation, storage facilities and • Seaports/airports • communications
Other rural infrastructure services

If the hypotheses implied by Figures 3 and 4 are correct, i.e., that the flows must lead toward a
virtuous cycle of localized (regional) linkages if national objectives of a more even pattern of
spatial development are to be achieved, then policy interventions would also be oriented toward
improving the chances for reciprocal benefits to accrue from the flows. These interventions
would not focus on the flows alone, but would also consider, for example, action in rural socio-
economic relations, production, institutional capacity, and environmental spheres.

As an example, a study of the flow of people might find that few go to local towns for
employment. Reasons for the low labor absorptive capacity of the town would then be identified
and policy options to improve its capacity would be considered. Or, research may find that
almost no information is provided from towns on market prices, crop prospects and other vital
areas of rural interest. Tracing the value added of a rural commodity from the farm to final
consumption would reveal not only how much of the total value added is captured by the region
of origin (usually quite low), but also the income implications of locally capturing greater shares
of downstream processing and distribution. Conversely, data may show that roads connecting a
village have enabled widespread income increases through greater flows of products to external
markets, which has led to more frequent visits to towns for shopping and farm inputs, with the
further result of spurring the growth of the town.

Of critical importance in any of these scenarios is not to assume that enhancing the physical
connectivity between rural and urban areas will bring either benefits or disbenefits to cities and
villages. Rather the task is to study each type of flow to discern the ways in which roads and
other infrastructure actually affect the nature and distribution of benefits from rural-urban

Finally, while research could be conducted between a single village and a single center, the
observation that rural-urban relations generally involve multiple linkages among villages and
with local as well as very distant cities suggests that the focus would be better placed in the
context of regional networks, with flows studied both from individual villages and among urban
centers in an identified regional cluster. In Indonesia, a convenient starting point for the research
would be to use a select number of kawasan andalan, which are prototype regional clusters
identified by the government for priority attention. By studying the dynamics of flows and
structural change within these clusters, policies can be better oriented to local conditions to
support Indonesia’s objectives of creating more symbiotic and reciprocal linkages between rural
and urban development.


Armstrong, W. and T.G. McGee (1985), Theaters of Accumulation (London: Methuen).

Birgegård, Lars-Erik (1988), “A Review of Experiences with Integrated Rural Development”,
Manchester Papers on Development, IV:1, Jan, 2-27.
Bunchorntavakul, Daranee (1976), The Impact of the Growth Center on Rural Areas: a Case
Study in the North of Thailand (Bangkok: AIT/HSD).
Chambers, Robert (1985), Rural Development: Putting the Last First (London: Longman)
Cohen, John M. (1980), “Integrated Rural Development: Clearing the Underbrush,” Sociologica
Ruralis, 20, 195-211.
Cohen, Monique, John English and Harold Brookfield (1977), “Functional Diversity at the Base
of the Urban System of Peninsular Malaysia,” J. Tropical Agriculture, 45, Dec., 12-25.
Collier, William, Kabul Santoso, Soentoro, Rudi Wibowo (1993), “A New Approach to Rural
Development in Java: Twenty-five Years of Village Studies” (Jakarta: PT Intersys Kelola
Maju for ILO).
de Jong, W. (1988),The Role of Towns in Rural Development: A Case Study of Banjarnegara,
Central Java (Yogyakarta: UGM Press).
de Jong, W., J.G.L. Palte and F. van Steenbergen (1983), “The Role of Banjarnegara as a Small
Regional Center in Central Java,” The Indonesian Journal of Geography, 13:45, June, 37-52.
Douglass, Mike (1981), “Thailand: Territorial Dissolution and Alternative Regional
Development for the Central Plains,” in W. Stöhr and D.R.F. Taylor, eds., Planning from
Above or Below? (Chichester: John Wiley), 183-208.
Douglass, Mike (1984), “National Urban Development Strategy Scenarios” (Jakarta: NUDS
Project, Ministry of Public Works, Tata Kota dan Tata Daerah).
Douglass, Mike (1990), “Regional Inequality and Regional Policy in Thailand: an International
Comparative Perspective,” (Bangkok: TDRI, Background Report 3-3, National Urban
Development Policy Study).
Douglass, Mike (1995), “Global Interdependence and Urbanization: Planning for the Bangkok
Mega-Urban Region,” in T.G. McGee and Ira Robinson, eds., The New Southeast Asia:
Managing the Mega-urban Regions (Vancouver: Univ. of British Columbia Press), 45-79.
Douglass, Mike (1997), “Structural Change and Urbanization in Indonesia: from the ‘Old’ to the
‘New’ International Division of Labor,” in Gavin Jones and Pravin Visaria, eds.,
Urbanization in Large Developing Countries; China, Indonesia, Brazil, and India (Oxford:
Clarendon Press), 111-141.
Fei, John C. and Gustav Ranis (1964), Development of the Labor Surplus Economy (Homewood,
Ill.: Irwin).
Friedmann, John (1968), “The Strategy of Deliberate Urbanization,” AIP Journal, Nov, 364-
Friedmann, John (1992), Empowerment: The Politics of Alternative Development (Oxford: Basil
Friedmann, John and Clyde Weaver (1979), Territory and Function (London: Edward Arnold).
Friedmann, John and Mike Douglass (1978), “Agropolitan Development: Toward a New
Strategy for Regional Planning in Asia,” in F. Lo and K. Salih, eds., Growth Pole Strategy
and Regional Development Policy (Oxford: Pergamon Press), 163-192.
Gibb, Arthur (1984), “Tertiary Urbanization: the Agricultural Market Center as a Consumption-
related Phenomenon,” Regional Development Dialogue, 5:1, Spring, 110-148.

Grandstaff, Somluckrat (1990), The Role of Demand in Provincial Industrialization (Bangkok:
Thailand Development Research Institute).
Hirschman, A. (1958), “Balanced Growth: A Critique,” The Strategy of Economic Development
(Yale), 50-75.
Johnson, E.A.J. (1970), “The Organization of Space in Developing Countries” (Cambridge,
Mass: Harvard University Press).
Khan: “The Comilla Model and the Integrated Rural Development Programme of Bangladesh --
An Experiment in ‘Cooperative Capitalism,’” in Dharam Ghai, Azizur Rahman Khan and
Eddy Lee, eds., Agrarian Systems and Rural Development, New York: Holmes and Meier,
1979), pp. 113-158.
Kammeier, D. and P. Swan, eds., (1984), Equity with Growth? Planning Perspectives for Small
Towns in Developing Countries (Bangkok: Asian Institute of Technology).
Khan, A. (1979), “The Comilla Project: and Experiment in Cooperative Capitalism,” World
Koppel, Bruce (1987), “Does Integrated Rural Development Really Work? Lessons from the
Bicol River Basin,” World Development, 15:2, 205-220.
Ladavalya, M.L. Bhansoon and Vasant Siripool (1988), “The Impact of the Regional City
Development Project upon Outlying Areas: the Chiang Mai Case” (Chiang Mai: SRI,
Chiang Mai University).
Lightfoot, P., Theodore Fuller and Peerasit Kamnuansilpa (1983), “Circulation and Interpersonal
Networks Linking Rural and Urban Areas: the Case of Roi-et, Northeastern Thailand”
(Honolulu: East-West Center, Papers of the East-West Population Institute ; no. 84).
Lipton, Michael (1977), Why Poor People Stay Poor (London: Gower Publishing Limited).
Mathur, Om, ed., (1982),The Role of Small Cities in National Development (Nagoya: UNCRD).
Maude, Alaric (1983), “Do Rural Towns Stimulate Rural Development?” Malaysian Journal of
Tropical Agriculture, 8, December, 40-48.
Myrdal, Gunnar (1957), Economic Theory and Underdeveloped Regions, London: Duckworth.
Richardson, Harry (1978), Regional and Urban Economics (London: Penguin).
Rondinelli, Dennis (1979), “Applied Policy Analysis for Integrated Regional Development
Planning in the Philippines,” Third World Planning Review, 1:2, 151-178.
Shiann-Far, Kung (1986), Growth Relation between Urban Center and Its Hinterland: A Case
Study of Surat Thani in Southern Thailand (Bangkok: AIT/HSD, Master’s Thesis).
Singer, Hans (1964), “The Generative versus Parasitic Role of Cities in Development,” in W.
Alonso and J. Friedmann, eds., Urban and Regional Planning (Boston, Mass: MIT Press).
Slater, David (1989), “Territorial Power and the Peripheral State: The Issue of
Decentralization,” Development and Change, 20:3, July, 501-531.
Williamson, J. (1965), “Regional Inequality and the Process of National Development: A
Description of the Patterns,” Economic Development and Cultural Change, 13:4 July, 3-45.